How to use it in python after TRT Build? #13905
Unanswered
davodogster
asked this question in
Other Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi I just did the TRT build using
./build.sh --cudnn_home /usr/local/cuda/ --cuda_home /usr/local/cuda-11.6 --use_tensorrt --tensorrt_home /home/scion.local/davidsos/Documents/onnxruntime/TensorRT-8.4.1.5
It passed most of the tests but failed the QOrdered tests.
But how does python recognize it when I run import onnxruntime?
I have a Pytorch lightning model I want to convert to ONNX and then to TRT for Onnxruntime-GPU TRT, CUDA Execution accelerated inference large scale (1M+ images). Is there a tutorial to convert a pytorch lightning model to ONNX-TRT then use it for TRT inference?
Thanks, Sam @abudup
Beta Was this translation helpful? Give feedback.
All reactions